Will you support Intel Arc?

[nathanodle]

I’m curious if you will support Arc, neural compressor would particularly benefit those platforms! Thanks!

[chensuyue]

We support general quantization/inference for Arc based on IPEX and ITEX. Here is an example for IPEX: https://github.com/intel/neural-compressor/blob/master/examples/pytorch/nlp/huggingface_models/question-answering/quantization/ptq_static/ipex/README.md#2-quantization-with-xpu
Example for ITEX: https://github.com/intel/neural-compressor/blob/master/examples/tensorflow/nlp/bert_base_mrpc/quantization/ptq/README.md#quantizing-the-model-on-intel-gpumandatory-to-install-itex https://github.com/intel/neural-compressor/blob/master/examples/tensorflow/nlp/bert_base_mrpc/quantization/ptq/README.md#quantization-config
Quantization steps are similar to CPU, just some config difference. The complex part is how to setup a XPU working ENV with IPEX/ITEX.

[nathanodle]

OK thanks. In that case you may want to update the readme?

Thanks for the response!

[thuang6]

No plan to claim formal support on ARC due to limited scope of validation. Close this issue for now.